Using Generative Models to Create a Visual Description of Climate Change
摘要
The discrepancy between the rapid dissemination of information and its effective communication underlies the phenomenon of scientific denialism. Given the rapid strides in AI generative models, this project explores the domain of knowledge visualization to portray weather data through a visually captivating representation of Rio de Janeiro’s climate evolution up to the year 2100. The use of prompt engineering over climate models has yielded promising outcomes in image generation, yet challenges remain in ensuring deterministic accuracy in image construction.